aboutsummaryrefslogtreecommitdiffstats
path: root/drivers/video
Commit message (Collapse)AuthorAgeFilesLines
* omap: dss: Suspend DSS with earlysuspendPaul Kocialkowski2017-05-111-0/+36
| | | | Signed-off-by: Paul Kocialkowski <contact@paulk.fr>
* Do not clear VRAM at initialization.Luden2016-04-031-0/+8
| | | | | Should help to avoid showing blank screen for extended periods of time during the boot.
* dss/manager: fix possible race by only touching cpr_coefsMichael Gernoth2016-01-241-25/+9
| | | | | | | | | If manager info got changed in between the get_info and the set_info_nocb in cpr_coef_store, this might lead to dropping important changes in the manager info. Fix this by just modifying the cpr_coefs in the existing info and not touching the rest. Change-Id: Iaab095e7532ff4f69c204e1c7322d64099957f84
* dss/manager: optimistically set cpr-valuesMichael Gernoth2016-01-241-5/+33
| | | | | | | | We are not concerned when we lose a few intermediate cpr-values because the user likes to play with the sliders. It's way better than crashing. Change-Id: Iaa90d4d5f1c87b71f06b420cf355b5b8d3899ca3
* dsscomp: don't overwrite cpr-settingsMichael Gernoth2016-01-241-2/+0
| | | | Change-Id: I6f8698139d8e0d539b04b6429e75a864ab38144a
* fs: push i_mutex and filemap_write_and_wait down into ->fsync() handlersJosef Bacik2015-05-021-2/+9
| | | | | | | | | | | | | | | | | | | | Btrfs needs to be able to control how filemap_write_and_wait_range() is called in fsync to make it less of a painful operation, so push down taking i_mutex and the calling of filemap_write_and_wait() down into the ->fsync() handlers. Some file systems can drop taking the i_mutex altogether it seems, like ext3 and ocfs2. For correctness sake I just pushed everything down in all cases to make sure that we keep the current behavior the same for everybody, and then each individual fs maintainer can make up their mind about what to do from there. Thanks, Acked-by: Jan Kara <jack@suse.cz> Signed-off-by: Josef Bacik <josef@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Conflicts: Documentation/filesystems/porting Change-Id: I7697244cce355e6d443a39c5133a5c94f02e199b
* OMAP4: HDCP: add support for M0/V' check from user space libsrinivas pulukuru2015-04-181-16/+2
| | | | | | | Add support for M0 and V' calculations from user space lib Change-Id: Ifb55d44723885d307c39d4bb7eb06cd4d50a469f Signed-off-by: srinivas pulukuru <srinivas.pulukuru@ti.com>
* OMAP: HDCP: Cancel pending start on disable/hpd dropMike J. Chen2015-04-181-5/+7
| | | | | | | | | | | | | | | | | | We observed a race condition where userland invoked an ioctl do blank/unblank the display while HDCP was pending to start (but hadn't yet started). The hdcp_irq_cb() wasn't cancelling that pending start. This caused the hdcp state machine to start when it shouldn't have been, interfering with the hpd state machine's reading of EDID once the unblank happened. The two state machines should really be unified at some point. Change-Id: Ief74d7eaa099f5455f6a42c65e98b965cdc50217 Signed-off-by: Mike J. Chen <mjchen@google.com> Signed-off-by: Ruslan Bilovol <ruslan.bilovol@ti.com>
* OMAPDSS: DSS: Scale DSS OPP according to pixel clock rateTaras Kondratiuk2015-04-181-0/+13
| | | | | | | | | | Before enabling pixel clock DSS should set its OPP for OPP framework to set correct VDD_CORE voltage level. When pixel clock is disabled DSS can set the lowest OPP. Change-Id: Ia1459614ae306dba9b048b149e474442e2880bba Signed-off-by: Taras Kondratiuk <taras@ti.com>
* tuna: dsi: fixes for tunaZiyan2015-04-181-0/+18
|
* OMAPDSS: DSI: add missing includeZiyan2015-04-181-0/+1
|
* OMAPDSS: DSI: fix debug messagesSergiy Kibrik2015-04-181-6/+6
| | | | | | | | Routine names in debug messages of dsi_vc_gen_read_2() and dsi_vc_dcs_read() are set to same "dsi_vc_rcs_read". Fix routine names in these messages. Change-Id: I9cceca5711932ba2c2ea3f2919baf461e30ddd8b Signed-off-by: Sergiy Kibrik <sergiy.kibrik@globallogic.com>
* OMAPDSS:DSI: dispose of static bpp definitionSergiy Kibrik2015-04-181-11/+8
| | | | | | | | | | | Calculation of bytes per pixel in dispc_to_dsi_clock() is not precise, based on static value definition. Instead, it should be calculated in-place, based on actual pixel size, which produce more accurate numbers in situations where remainder appears (e.g. 18-bit pixel size). Change-Id: I4c90b2ec5e1d6139d84cbdc55af3f2a484785968 Signed-off-by: Sergiy Kibrik <sergiy.kibrik@globallogic.com>
* OMAP4: DSI: Reset required disable-enable sequenceSujeet Baranwal2015-04-181-0/+6
| | | | | | | | | | During the initialization of DSI,some times BTA sent by OMAP DSI, was not receiving an BTA ACK. To rectify this, we need to reset the DSI module whenever we diable-enable the DSI module. Change-Id: Ic2542970da06a121559a8cc700222881c60c025c Signed-off-by: Sujeet Kumar Baranwal <s-baranwal@ti.com>
* OMAP4: DSI: DSI reset removal during enablingSujeet Baranwal2015-04-181-22/+19
| | | | | | | | | | | | | | | | DSI module is made wake up capable in its HWMOD flags but in The function omapdss_dsi_display_enable(), DSI module is reset thus values set in DSI_SYSCONFIG is lost because of which DSI is no longer wakeup capable. This prevents any interrupt from DSI to be permeated to A9 when A9 sleeping.this reset is not required while enabling the display thus now it is removed.so it retains the HWMOD settings. Now with this fix, the mdelay() within dsi_vc_send_bta_sync(), is removed and a proper interrupt based implementation is done. Change-Id: If6c3d25a5ba8003a0bd7732af0fcd81a839293aa Signed-off-by: Sujeet Kumar Baranwal <s-baranwal@ti.com>
* OMAP4: DSS: Video mode: Fixed dsi horizontal blankings calculation.Jorge Bustamante2015-04-181-52/+21
| | | | | | | | | | | | | | | | | | | | | | | | | | | Fixed calculation for tl, hsa, hbp and hfp parameters used on dsi video mode. dispc_to_dsi() is simply converting from PCLK to TXBYTECLK. TXBYTECLK is CLKIN4DDR / 16, while CLKIN4DDR should be: (PCLK * bytes_per_pixel * 8 * 4) / (lanes * 2), so TXBYTECLK is simply (PCLK * bytes_per_pixel) / lanes. hsa is not used by dsi engine so it is added to hbp. tl should be an exact integer, so it is protected by a WARN on case it isn't. Removed unnecesary configuration on dsi_videomode_panel_preinit(). Tested with 1024x768, 1024x600, 1280x800, 800x600, 960x960 resolutions and different combinations of horizontal blankings. (tablet1 & tablet2 cases included). Hasn't been tested with 1 or 2 data lane configurations or with a different pixel format than RGB888. Change-Id: Idc46d1d316febed947fe0c72c5060c74ff40d6de Signed-off-by: Jorge Bustamante <jbustamante@ti.com>
* [OMAP4-DSS-Tablet2] Add parameter to fine-tune the DSI clock timing registerSreenidhi Koti Ananda Rao2015-04-181-2/+6
| | | | | | | | | | | | | | | Offset the values of the DDR_CLK_PRE and DDR_CLK_POST fields based on panel requirements. Note: Blaze tablet panel does not requires this offset, hence setting it to 0. The DDR_CLK_ALWAYS_ON bit & BLANKING_MODE bits should be set based on the updated value (non-zero) of the ddr_offset. Change-Id: I1715c293c5ec46a4765293867bf62efd58b11b6a Signed-off-by: Gilles-Arnaud Bleu-Laine <gilles@ti.com> Signed-off-by: Sreenidhi Koti <sreenidhi@ti.com>
* [Tablet] Add identifier for Tablet 1.2Dan Murphy2015-04-181-1/+2
| | | | | | | | | | | | There are 3 tablet 1 versions circulating. Need to add the third identifier so that the display boots properly Change-Id: I16b298d537abe771ff9ac4e3b4fa90621522b510 Signed-off-by: Dan Murphy <dmurphy@ti.com> Conflicts: arch/arm/mach-omap2/board-44xx-tablet-panel.c arch/arm/plat-omap/include/plat/omap_apps_brd_id.h
* [Blaze] DSI: Fix the HACK for the OMAP DSI for BLAZEDan Murphy2015-04-181-1/+1
| | | | | | | | | When the apps board ID values changed the DSI HACK in place for video vs non-video mode broke so we were sending some OCP_WIDTH and the DSI was returning errors` Change-Id: Icb1034ff244ed7ddaa877dcf7b39f94c642eff6a Signed-off-by: Dan Murphy <dmurphy@ti.com>
* [Tablet] Modify board ID for tablet 2.1Dan Murphy2015-04-181-1/+2
| | | | | | | | | | | | | | Add tablet 2.1 ID and 2.1.1 ID so we can differentiate between the different apps boards Change-Id: I50e8ad9cdc6c03b100e353588adb532fa99b43d0 Signed-off-by: Dan Murphy <dmurphy@ti.com> Conflicts: arch/arm/mach-omap2/board-44xx-identity.c arch/arm/mach-omap2/board-44xx-tablet-panel.c arch/arm/mach-omap2/omap4_apps_brd_id.c arch/arm/plat-omap/include/plat/omap_apps_brd_id.h
* [Tablet] HACK: This is yet another HACK for the displayDan Murphy2015-04-182-4/+16
| | | | | | | | | | | | | This is a hack of the dsi for the tablet and tablet2. Basically this patch will allow tablet and tablet2 to boot from the same kernel and display appropriately. The permenant fix needs to be found. Signed-off-by: Dan Murphy <dmurphy@ti.com> Conflicts: arch/arm/mach-omap2/board-44xx-tablet-panel.c
* [DSI] HACK: Add check for blaze to not set second OCP WidthDan Murphy2015-04-181-3/+11
| | | | | | | | | This is a HACK as performing this command on blaze causes DSI errors and does not allow blaze to display anything for now cause it to skip on blaze but allow this on tablet video displays Signed-off-by: Dan Murphy <dmurphy@ti.com>
* [Tablet] DSI Changes for tablet2Tomi Valkeinen2015-04-181-52/+299
| | | | | | | Modify the DSI layer for the tablet2 panel Change-Id: I00b430755c1776819de773e389b7d2814c24dd01 Signed-off-by: Dan Murphy <dmurphy@ti.com>
* [Blaze] HACK:OMAP4: DSI Command mode fix for BlazeDandawate Saket2015-04-181-0/+3
| | | | | | | | | Video mode support fix for channel 0 32bit alignement was causing command mode to break. Need more investigation on the video mode platform to check why this break is seen. Change-Id: I0c4a82a723e017e1c90016807117d2dcd2c2453e Signed-off-by: Dandawate Saket <dsaket@ti.com>
* OMAPDSS: HDMI: report HDMI as manually updated panelMuralidhar Dixit2015-04-171-0/+7
| | | | | | | | | | Report HDMI as manually updated panel. Currently it is default treated as auto updated panel by OMAPFB. This is used by OMAPFB to determine if VSYNC notification is possible by panel or not. Change-Id: I3d1fe91f79fd2a39497bea700f9653ef141b310f Signed-off-by: Muralidhar Dixit <murali.dixit@ti.com>
* OMAP4: HDMI: Detect HDMI connection state at probe timeSujeet Baranwal2015-04-171-0/+3
| | | | | | | | Check if HDMI is connected at probe time, because we will not get an HPD interrupt in this case. Change-Id: I0bb3a25773edfef787f14cccb55a9bc231cde025 Signed-off-by: Baranwal Sujeet <s-baranwal@ti.com>
* OMAPDSS: CEC: Physical address discovery form edidMuralidhar Dixit2015-04-171-1/+19
| | | | | | | | HDMI sink device will set the physical address for CEC devices Change-Id: I6419834816978dfc8bcb8ef9a7b6c36276e5792c Signed-off-by: Muralidhar Dixit <murali.dixit@ti.com>
* OMAP: HDMI: Don't release power constraints which were never added.John Grossman2015-04-171-4/+11
| | | | | | | | | | Make certain we don't call pm_qos_remove_request if we have never added the request in the first place as it produces a nasty looking error message and stack track in the kernel logs. Change-Id: I1e955f15944908706f3a67cf6be0603478036ea5 Signed-off-by: John Grossman <johngro@google.com> Signed-off-by: Ruslan Bilovol <ruslan.bilovol@ti.com>
* OMAP4: HDMI: HACK: Set c-sate constraint for HDMIsrinivas pulukuru2015-04-171-0/+25
| | | | | | | | | | | This is a temporary HACK and needs to be rootcaused and fixed. Set PM c-state constraints when HDMI is connected so that core doesn't hit below C2. This will keep the system from locking up when HDMI is connected. Change-Id: I341922f3d890b416a78fe6a5aabf0860c6b212ea Signed-off-by: Srinivas Pulukuru <srinivas.pulukuru@ti.com>
* [HDMI] Add HMDI S3D supportAlberto Aguirre2015-04-175-3/+173
| | | | | | | | | | | | Add Video specific infoframe configuration. Add sysfs interface to change s3d type and disable/enable s3d mode. Change-Id: I8dc6621a5d4fab79ba95ecbc8e23ade9b6c5dfab Signed-off-by: Alberto Aguirre <a-aguirre@ti.com> Conflicts: include/video/hdmi_ti_4xxx_ip.h
* OMAPDSS: CEC: changing the cec switch nameMuralidhar Dixit2015-04-171-1/+1
| | | | | | | | Changing the CEC event to avoid conflict with HDMI event Change-Id: I08748e6a7b2879ffec12f009afb8b03522247309 Signed-off-by: Muralidhar Dixit <murali.dixit@ti.com>
* [HDMI] Expose HDMI EDID block through sysfsAlberto Aguirre2015-04-173-3/+28
| | | | | | | Expose raw HDMI EDID binary data through sysfs. Change-Id: I8ae8b9cad202d0bccb395725b4a74bbdb43ee514 Signed-off-by: Alberto Aguirre <a-aguirre@ti.com>
* Input: CEC keyboard driverMuralidhar Dixit2015-04-172-1/+85
| | | | | | | | | Adding support for CEC key board driver in input device. UI commands received from CEC are mapped to the key buttons in the linux kernel. Change-Id: If4136940f28fb4a0c4b9b730aabb8e0fc7837ccf Signed-off-by: Muralidhar Dixit <murali.dixit@ti.com>
* OMAP4: HDMI: Fix HDMI_TXPHY_TX_CONTROL.FREQOUT programmingVaradarajan, Charulatha2015-04-172-3/+11
| | | | | | | | | | | According to TRM, the HDMI_TXPHY_TX_CONTROL.FREQOUT bit has to be programmed based on the clock frequency. But this bit is always programmed as 1 in the HDMI driver irrespective of the clock frequency values. Fix this and program HDMI_TXPHY_TX_CONTROL.FREQOUT based on the clock frequency. Change-Id: I25a4c335ec773e241cc0daca5c1c1c8f223cd4a0 Signed-off-by: TAKAHASHI, Masato <m_takahashi@ti.com> Signed-off-by: Varadarajan, Charulatha <charu@ti.com>
* OMAPDSS: CEC: Adding HDMI CEC supportMuralidhar Dixit2015-04-172-1/+455
| | | | | | | Adding CEC driver support in HDMI Change-Id: Ia6a7d04b1df3a459ae435f8bec79219b2036dfa0 Signed-off-by: Muralidhar Dixit <murali.dixit@ti.com>
* OMAPDSS: CEC: Adding support for CEC in HDMI driverMuralidhar Dixit2015-04-172-0/+459
| | | | | | | | | | | Adding support for CEC functionality in OMAP4 HDMI IP specific file Change-Id: I333e8ae484e0e56e0c597bca754f0276614d90ee Signed-off-by: Muralidhar Dixit <murali.dixit@ti.com> Conflicts: include/video/hdmi_ti_4xxx_ip.h
* OMAPDSS: CEC: Add callback to CEC driver on HDMI HPDMuralidhar Dixit2015-04-175-2/+60
| | | | | | | | Adding support for call back from HDMI to CEC driver on HPD and HDMI enable. Change-Id: I9fff9d41d598e5ef375bc03faafb9a559ce7435a Signed-off-by: Muralidhar Dixit <murali.dixit@ti.com>
* added checks for HDMI enabledHashcode2015-04-171-0/+2
|
* dss: update variable names to reflect kernel changesZiyan2015-01-131-2/+2
| | | | Change-Id: I4c66e9825dce4e41fc11477ee40af5e16191db5f
* DSS: Fixed compilation warnings at device.c:504Ievgen Musiichuk2015-01-131-1/+2
| | | | | | | | | | | Warning: comparison of distinct pointer types lacks a cast Changed clamp macro to clamp_t with __u32 type Change-Id: I1ab6169a79e8dfdfb2e3b87147e225382a64248d Signed-off-by: Ievgen Musiichuk <ievgen.musiichuk@ti.com> Conflicts: drivers/video/omap2/dsscomp/device.c
* Revert "Revert "DSSCOMP: Added UV buffer user-space address to the API""Pavel Nedev2015-01-131-2/+6
| | | | | | | | | | This reverts commit 2ca61536d413af33d9ef58db249a327991f62a79. Change-Id: I9d00daf95698b14d5029a49f3802400e8f60d4da Signed-off-by: Pavel Nedev <pnedev@mm-sol.com> Conflicts: include/video/dsscomp.h
* Revert "OMAP: DSS2: Support for Samsung AMS452GN05"Ziyan2015-01-121-25/+0
| | | | This reverts commit 968b17621a48a4ea5049968d3601f59624f33dd3.
* OMAP: DSS: Add device latency constraintGilles-Arnaud Bleu-Laine2015-01-121-0/+10
| | | | | | | | | | | | Request a latency constraint on PM framework for DSS before cutting module clocks. Release the request upon enabling the clocks. Change-Id: If257218a3cb8c3ea0d8451421f87e0a9ba199bb5 Signed-off-by: Gilles-Arnaud Bleu-Laine <gilles@ti.com> Conflicts: drivers/video/omap2/dss/dispc.c
* OMAP: DSS: add support DPI panel for Panda in waitforvsyncVince Kim2015-01-121-12/+42
| | | | | | | | | | | | Panda board uses DPI interace for DVI, and vsync IRQ for DPI panel is added to avoid timeout. As adding DPI pannel support, if...else statement is re-written with switch statement. Other non-supported panels are added in switch statement as place holer. Change-Id: Ic383acd55245e86579a29bb04cfd50334df0e133 Signed-off-by: Vince Kim <vince.kim@ti.com>
* OMAP4: DSS: Fix to consider DSI manual update mode in waitforvsyncSunita Nadampalli2015-01-121-3/+11
| | | | | | | | | | | Add checks for manual update mode of DSI, and wait for framedone interrupt in case of manual update mode. Change-Id: I143a8d457219e2dab59ef099bab5c31043a36da0 Signed-off-by: Sunita Nadampalli <sunitan@ti.com> Conflicts: drivers/video/omap2/dss/manager.c
* Added Color Control version 4.Ezekeel2014-12-161-0/+40
| | | | | | | Conflicts: drivers/misc/Makefile Change-Id: Ie57d33003d1ce0a750ef2ec2cb15884a449df92a
* Merge remote-tracking branch 'linux/linux-3.0.y' into stable-newpvrZiyann2014-11-262-29/+26
|\ | | | | | | | | Conflicts: arch/arm/include/asm/hardware/cache-l2x0.h
| * fbcon: when font is freed, clear also vc_font.dataMika Kuoppala2013-05-071-0/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | commit e6637d5427d2af9f3f33b95447bfc5347e5ccd85 upstream. commit ae1287865f5361fa138d4d3b1b6277908b54eac9 Author: Dave Airlie <airlied@redhat.com> Date: Thu Jan 24 16:12:41 2013 +1000 fbcon: don't lose the console font across generic->chip driver switch uses a pointer in vc->vc_font.data to load font into the new driver. However if the font is actually freed, we need to clear the data so that we don't reload font from dangling pointer. Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=892340 Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com> Signed-off-by: Dave Airlie <airlied@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
| * vm: convert fb_mmap to vm_iomap_memory() helperLinus Torvalds2013-04-251-26/+14
| | | | | | | | | | | | | | | | | | | | | | | | | | | | commit fc9bbca8f650e5f738af8806317c0a041a48ae4a upstream. This is my example conversion of a few existing mmap users. The fb_mmap() case is a good example because it is a bit more complicated than some: fb_mmap() mmaps one of two different memory areas depending on the page offset of the mmap (but happily there is never any mixing of the two, so the helper function still works). Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
| * fbcon: fix locking harderDave Airlie2013-04-252-3/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | commit 054430e773c9a1e26f38e30156eff02dedfffc17 upstream. Okay so Alan's patch handled the case where there was no registered fbcon, however the other path entered in set_con2fb_map pit. In there we called fbcon_takeover, but we also took the console lock in a couple of places. So push the console lock out to the callers of set_con2fb_map, this means fbmem and switcheroo needed to take the lock around the fb notifier entry points that lead to this. This should fix the efifb regression seen by Maarten. Tested-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> Tested-by: Lu Hua <huax.lu@intel.com> Signed-off-by: Dave Airlie <airlied@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
generated by cgit v1.1 at 2025-01-13 14:35:03 +0000